Speed is not a technical detail but a business figure. Every second a page loads earlier measurably shifts how many visitors turn into buyers and how much they put in the cart. The hard numbers are unambiguous: a delay of just 100 milliseconds can cut conversions by 7 percent (Akamai, 2017), while improving mobile load time by only 0.1 seconds raised retail conversions by 8.4 percent (Deloitte, 2020). And patience is thin: 53 percent (Google, 2016) of mobile visits are abandoned when loading takes longer than three seconds. This article works out the revenue lever of speed -- per 100 milliseconds, per second -- and translates your Core Web Vitals into euros.
Key takeaways
- Load time is a revenue figure: depending on the starting point, every second saved visibly lifts conversion -- and even 100 milliseconds have a say.
- Improving mobile load time by 0.1 seconds raised retail conversions by 8.4 percent and cart value by 9.2 percent (Deloitte).
- Mobile bounce starts early: beyond three seconds more than half of visitors leave, and the probability of a bounce rises sharply with every further second.
- A shop under two seconds converts markedly better than one at three to four seconds -- the decisive threshold sits in the first visible area.
- Core Web Vitals are not an end in themselves: they can be translated into revenue lost or gained, becoming the basis of a sound investment decision.
Why Speed Acts Directly on Revenue
The link between load time and revenue is not a guess but documented across large datasets. Together with SOASTA, Google analysed behaviour across millions of mobile sessions and found a clear effect: as load time rises from one to three seconds, the probability of a bounce increases by 32 percent (Google/SOASTA, 2017). Every additional second is therefore not neutral but costs part of the audience before they have even seen a product.
The mechanics behind it are simple. A slow page creates waiting, and waiting creates abandonment. Whoever bounces sees no products, adds nothing to the cart and places no order. The revenue loss thus arises not at the end of the buying process but right at the start -- at the first interaction with the page. That is exactly why an improvement in load time acts on every following step: more visitors stay, more see products, more reach the cart. This chain can be traced for your own shop with a clean performance analysis.
Why the First Seconds Decide the Revenue
The Revenue Lever per 100 Milliseconds
How fine-grained the link is shows even at the millisecond level. According to Akamai's analysis, a delay of just 100 milliseconds can lower conversions by 7 percent (Akamai, 2017) -- a figure that looks small at first glance but quickly adds up to substantial amounts in a high-revenue shop. The data basis was extensive: the analysis drew on around 10 billion (Akamai, 2017) anonymous user visits to leading online retailers.
Seen the other way round, speed is a growth lever. In the Google-commissioned Deloitte study titled Milliseconds Make Millions, reducing mobile load time by just 0.1 seconds (Deloitte, 2020) improved several metrics at once. In retail, conversions rose by 8.4 percent (Deloitte, 2020) and average cart value by 9.2 percent (Deloitte, 2020). In addition, page views per session grew by 2 percent (Deloitte, 2020). The study observed more than 30 million user sessions and thus makes visible the effect that individual milliseconds have on real purchase decisions.
More Conversions
Just 0.1 seconds faster mobile load time raised retail conversions by 8.4 percent (Deloitte). Speed turns existing traffic into more orders without a single additional visitor being bought in.
Higher Cart Value
Over the same period, average cart value rose by 9.2 percent (Deloitte). A fluid page lowers the barrier to viewing and adding more items -- revenue per order grows with it.
More Product Views
Page views per session grew by 2 percent (Deloitte). Whoever navigates quickly through categories and products sees more offers -- and more products seen means more opportunities to buy.
The Mobile Bounce Beyond Three Seconds
On the phone, tolerance is especially low. 53 percent (Google, 2016) of mobile visits are abandoned when loading takes longer than three seconds -- more than half, before the content has fully appeared. Because the mobile channel carries the largest part of traffic, this bounce is no edge case but hits the bulk of visitors. How to measure the mobile channel cleanly in lab and field data is shown in our article on RUM and CrUX field data.
The probability of a bounce does not rise linearly but accelerates. As load time grows from one to three seconds, the probability of a bounce increases by 32 percent (Google/SOASTA, 2017); at five seconds it is already 90 percent (Google/SOASTA, 2017). Every further second therefore costs a disproportionate number of visitors. Which technical conditions slow the mobile channel is explored in our article on mobile performance optimization.
| Load time | Bounce probability (relative) | Business consequence |
|---|---|---|
| 1 second | reference | highest yield of the visitor stream |
| 3 seconds | plus 32 percent (Google/SOASTA) | noticeable loss before the first product view |
| 5 seconds | plus 90 percent (Google/SOASTA) | a large part of visitors is lost |
| over 3 seconds (mobile) | 53 percent abandon (Google) | the main channel loses the majority of sessions |
The Average Hides the Mobile Loss
Core Web Vitals Translated into Euros
The Core Web Vitals -- Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift -- sound at first like pure engineering. Their business value only becomes tangible once they are converted into revenue. Global online retail now turns over around 6.4 trillion (eMarketer, 2025) US dollars, making up about 20.5 percent (eMarketer, 2025) of total retail. Around 57 percent (Statista, 2024) of this revenue happens on mobile -- exactly where load time most sharply decides between staying and bouncing.
For the individual shop, the lever can be illustrated with a simple worked example. It is an illustration based on the cited studies, not a promised result -- the actual effect depends on the starting point, assortment and audience:
- Starting point: a shop generates 500,000 euros in revenue per month at a conversion rate of 2 percent.
- Speed gain: mobile load time is improved by several tenths of a second and then sits stably under two seconds.
- Effect on conversion: applying the order of magnitude documented by Deloitte, an improvement in the single-digit percentage range can arise -- assume around 5 percent relative.
- Result: 500,000 euros would become roughly 525,000 euros per month on paper, without buying in a single additional visitor.
- Extra effect: if cart value rises in parallel, as observed in the study, the revenue gain is higher still.
Performance optimization is not a cost centre but an investment with measurable return. Whoever wins milliseconds wins orders -- from the traffic that has already been paid for.
This calculation makes clear why investing in good Core Web Vitals pays off twice: it lowers the bounce rate in the most important channel and at the same time raises organic visibility, because Google factors the user experience into its assessment. How to improve the three values concretely is shown in our overview of the Core Web Vitals in 2026; the link between sustainability and speed is deepened in the article on sustainable websites and their carbon footprint.
Under Two Seconds: the Threshold That Converts
A practical finish line can be derived from the data: a shop that loads noticeably under two seconds converts markedly better than one that takes three to four seconds. The reason lies in the accelerating bounce curve -- below two seconds the vast majority of visitors stay, above three seconds behaviour tips over. How critical the threshold is shows in another Akamai figure: a delay of just two seconds raised the bounce rate by 103 percent (Akamai, 2017), more than a doubling.
The path under the two-second mark rarely runs through a single measure but through the interplay of server and frontend. The server response time sets the lower bound from which anything can load at all; how to shorten it is covered in our article on TTFB and server response time. In the frontend, what matters most is how quickly the first visible content appears. The key levers at a glance:
- Lower the server response time so the browser can start rendering early
- Serve the largest visible element -- usually an image -- early and in a fitting size
- Convert images to modern, lighter formats and scale them per screen size
- Defer non-critical JavaScript so the page responds and becomes interactive sooner
- Avoid layout shifts so no mis-taps and no renewed waiting occur
- Measure every change against real field data, not just in the lab on a fast device
An often underestimated point is the order of loading. If the browser requests the most important image too late, the entire visible build-up is delayed. How to additionally speed up the switch between pages is shown in our article on Speculation Rules and instant navigation. Together, server optimization and a lean frontend ensure the shop falls below the decisive threshold -- a field that our server optimization for short response times works on deliberately.
From the Millisecond Gain to the Order
The business case for speed is thus complete: faster pages lower the bounce rate, raise conversion and increase cart value -- across all sectors. The Deloitte analysis found the effect not only in retail but also in travel, where conversions rose by 10.1 percent (Deloitte, 2020) at 0.1 seconds faster load time. Speed is therefore no niche topic for technology enthusiasts but a direct revenue driver. How broadly our range of services addresses this is shown in the overview of our performance services.
The Core in One Sentence
The first step is always measurement: only once the lost revenue is quantified can the investment in speed be soundly justified. That is exactly what a specialized performance analysis for e-commerce delivers -- it quantifies how much revenue the current load time costs and turns it into a sound priority list. Whoever improves the three metrics consistently, as our article on INP and the response time of the page shows, converts milliseconds into orders -- from traffic that is already there.
Sources and Studies